An Error Reduction Approach for Discrete Robust Train Tracking Control

Automatic train driving control is widely studied. Considering the uncertain disturbance during driving, robustness is fundamental in controller designs. This paper presents a method to design a robust controller, which guarantees the control error is bounded in an invariant ellipsoid range. An optimization algorithm is proposed to reduce the size of the ellipsoid. Experimental results illustrate that the size of optimized ellipsoid is smaller than non-optimal ones. An inter-station driving simulation shows that the train tracks the desired reference well. This method provides possibilities to further improve performance of railways and some potential application scenarios are also discussed here.